Identifying research contributions based on semantic analysis of citation sentences: A case study of the 2021 Physiology or Medicine Nobel Prize laureates

Prior research in identifying research contributions has extracted contributions based on the academic paper itself. In this study, we aim to explore the research contributions from the perspective of the external citation environment (citation sentences in this article), with the attainment of two 2021 Physiology or Medicine Nobel Prize laureates and their academic papers used as an experimental case. First, we analyze the features of texts indicating research contributions in academic papers, and then construct a model to identify citation sentences describing the significance of contributions in all the citing literature. In addition, we make a sentence-level contribution analysis of the important information contained in the contribution sentences. Finally, the clustering method is utilized to group similar contribution sentences into research contribution points, which allow us to automatically summarize the research contributions of the two laureates and compare these concluded contribution points to the award ceremony speech for validation. The set of approaches suggested in this work for identifying research contributions in academic papers could contribute significant semantic assessment data to reforming and rebuilding the research evaluation system.